Deep Learning based Improved Strategy for Credit Card Fraud Detection using Linear Regression

Abstract

Credit card fraud detection is a critical challenge in the financial industry, necessitating robust and accurate methods to identify fraudulent transactions. Traditional machine learning approaches have demonstrated some success but often struggle with the dynamic and evolving nature of fraudulent activities. This paper proposes an improved strategy for credit card fraud detection by integrating deep learning techniques with linear regression models.The proposed method leverages the strengths of deep learning in capturing complex, non-linear relationships and high-dimensional patterns within transaction data, while utilizing linear regression to ensure interpretability and simplicity in the final decision-making process. Our hybrid model first employs a deep learning architecture, specifically a convolutional neural network (CNN) or a recurrent neural network (RNN), to extract meaningful features from raw transaction data. These features are then fed into a linear regression model that performs the final classification.The integration of deep learning and linear regression not only boosts performance but also provides insights into the contributing factors of fraudulent transactions, aiding financial institutions in their ongoing efforts to combat credit card fraud.